DR. TYNDALL ON MOLECULAR INFLUENCES. 
229 
a cube of rock-crystal (pure silica) of the same size produces a deflection of 
90 °. 
This single experiment is suflScient to show how different must be the meteoro- 
logical effects of these two substances, when they exist in sufficient quantity to 
exercise an influence upon climate. Among the more prominent influences here, 
Humboldt mentions the nature of the soil and of vegetation. The general influence of 
an arid and exposed soil has been long known, but the part played by this substance, 
silica, has hitherto had no particular importance attached to it. Were gypsum, 
however, instead of silica, the prevalent mineral in Sahara, a very different state of 
things from the present would assuredly exist. A cube of the latter substance 
examined in the usual manner produces a deflection of 
19 ° 
only. It is scarcely superior to wood, while there is the strongest experimental 
grounds for the belief that silica possesses a higher conductive power than some of 
the metals. These grounds shall be adduced in a future paper. 
Let us consider, for a moment, the process which takes place from sunrise to the 
hour of maximum temperature in a region overspread with forests, and compare it 
with that which must take place in the African Desert. In the former case, the heat 
slowly and with difficulty penetrates the masses of wood and leaves on which it falls, 
and after the point of maximum temperature is passed, the yielding up of the heat 
acquired is proportionately slow. In the desert, however, the mass of silica exposed 
to the sun becomes burning hot as the hour of maximum temperature approaches ; 
but, after this is passed, the heat is yielded up with proportionate facility. Hence a 
maximum and minimum thermometer must, in the latter case, mark a far wider range 
of temperature during the twenty-four hours than in the former. This agrees with 
observation. In Sahara, to use the words of Mrs. Somerville, during “ the glare of 
noon the air quivers with the heat reflected from the red sand, and in the night it is 
chilled under a clear sky sparkling with its host of stars*.’ Were gypsum, however, 
the prevailing mineral, it is a priori certain that this could not be the case to anything 
like its present extent. 
The following experiments furnish some notion of the transmissive power of a few 
other organic structures : cubes of the substances were examined in the usual manner. 
Tooth of Walrus 16 
Tusk of East Indian Elephant . 17 
Whalebone 9 
Rhinoceros’s-horn 9 
Cow’s-horn 9 
Considering the density and elasticity of ivory, we might be disposed to attribute to 
it a comparatively high conductive power ; but the experiment proves it to be a very 
* Physical Geography, vol. i. p. 147. 
2 H 2 
